Abstract Two commonly used silvicultural herbicides (triclopyr and glyphosate) were examined for their effects on lichens in northeastern Ontario. One hundred 1 m 2 plots were randomly established throughout a single, open canopy forest stand with tree cover dominated by Picea mariana and Pinus banksiana and ground cover dominated by lichens. Herbicides were applied to the plots along a gradient of concentrations ranging from 0.71 to 6.72 kg acid equivalent ha −1 . The pre- and 1 year post-application abundance (percent cover) of 25 lichen species in 1 m 2 plots were compared. Triclopyr and glyphosate reduced the abundance of 40% and 56% of the lichen species studied, respectively. Tolerance to these herbicides varied among lichen species and a cluster analysis was used to define four tolerance classes. Analysis of variance was used to compare lichen abundance among the tolerance classes and multivariate analysis (canonical correspondence) was used to explore variation in lichen abundance as constrained by species traits post-herbicide application. Except for Trapeliopsis granulosa , the most herbicide sensitive species were richly branched. The species with the highest mortality were Bryoria furcellata , Cladonia uncialis , and T. granulosa . In general, lichen response to herbicide treatments cannot be assessed using ecological units; species must be considered individually or, as proposed here, by tolerance class. Study results provide forest managers with improved understanding of the effects of herbicide applications on non-target organisms.